Converting outputs to a new coordinate system using Convert feature

[aggieair]

Hello,

I am wanting to convert our outputs (ortho, DSM) into a new projection system using the tips on this HelpDesk page: https://agisoft.freshdesk.com/support/solutions/articles/31000151182-conversion-between-the-coordinate-systems

Our Agisoft set projection is WGS 84 / UTM Zone 12N + WGS84 Ellipsoid HT. This is LiDAR data it it captured in this projection.

However, we want our DSM and ortho - without ground control points (like we normally do) - converted into NAD83 (2011) UTM Zone 12N + NAVD88 Geoid 12B.

(Normally our GCP are in the desired projection and that is how we get our end projects in our desire projection, but this LiDAR doesn't use GCP.)

Having just discovered the Convert button in the Reference menu, I'm trying to make it work. It seems to convert the XY to the new projection, but not the Z so looks like we need to define things further if we want the Z to change too.

Can anyone point me in the right direction? Would I do this in the Console window instead? Any guidance on that would be appreciated.

[Alexey Pasumansky]

Hello aggieair,

All three XYZ coordinates should be transformed when Convert Reference command is used. Make sure that you have GeoTIFF file corresponding to NAVD88 Geoid 12B in /geoids/ subfolder of Metashape Pro installation directory.

[aggieair]

I have the Geoid 12B TIF in the C: directory, we use it often and navigate to the TIF in the Datum Transformation Parameters window

In the Datum Transformation Parameters window, it only lets me select:

CUSTOM
NONE
NAD83(2011) to WGS84 (1) (EPSG::9974)

The latter seems the closest, but how do I get it to say the opposite? From WGS84 (1) (EPSG::9974) to NAD83(2011) ?

[Paulo]

Hllo aggieair,

you just need to enter or use the correct  7 transformation parameters from NAD83(2011) to WGS84. The software will automatically use the correct set according to the direction of transformation...

The  latest set of parameters from NGS site are:

   TOWGS84                  
Epoch   TX(m)   TY(m)   TZ(m)   RX(as)            RY(as)           RZ(as)          DS(ppm)
2010   -1.0053   1.90921   0.54157   0.0267815   -0.00042078   0.01093254   -0.00036891

so when converting. enter these parameters in the TOWGS84 section of datum transformation as in:

[aggieair]

@Paulo, thanks for the additional tips regarding the seven transformation parameters.

I want to go from WGS84 to NAD83(2011) NAVD88 Geoid 12B. How do I get parameters for the opposite direction, starting in WGS84?

Now I'm trying to find where I can look up the seven parameters myself. I'm searching all sorts of words on National Geodetic Survey and Google and it's not clear. Hoping there is some king of user interface it can populate these values for me as I have more projections/parameter needs.

[Paulo]

Hi aggieair,

Just need to  enter these values once in the TOWGS84 section. So if you are doing WGS84 to NAD83(2011), the program will just invert the sign of the parameters. Strictly speaking, the above set is for epoch (decimal year) 2010, So if you want a rigorous set of parameters  for a survey in beginning 2025 (epoch 2025.1) you would have to take into consideration the parameter velocities as in following attachment...(Earth reference is dynamic)

For reference see following page from NGS https://www.ngs.noaa.gov/CORS/news/historical_helmert.shtml. It seems it stopped support fot these in 2017 so I am taking the parameters from NAD83(Canadian Spatial Reference System) to WGS84 which are basically the same as those for US.

So these parameters will apply to all NAD83(2011) CRS (projected or geographic).

[aggieair]

Thanks for sharing the Epoch 2010 and 2025.1, Paulo.

So I did a convert test using 3 variations:

1. "NAD83(2011) to WGS84" in the Custom drop down menu, plus hand selecting the us_noaa_g2012b.tif from my Geoids folder (name does not stick in window, but the console window recognizes it)

2. Parameters from Epoch 2012

3. Parameters from Epoch 2025.1

I export the DSM after each convert test and view them in ArcGIS.

The elevations are all identical, as if changing the parameters did not do anything to the elevations. I guess I was expecting some slight variations?

Meanwhile, the elevations I'm getting are in the right ballpark, but are still all off 0.6 meters from my known benchmarks.

When the DSM was still in WGS84 UTM (pre convert), my elevations were only 0.05 m off from my known benchmarks.

[Paulo]

It is strange. Normally the NAVD88 elevations should be a little different for each epoch as the h (elipsoidal height) changes a bit.

here are the coordinates of a point in WGS84 and then in NAD83(2011) + NAVD88 epoch 2010 and in NAD83(2011) + NAVD88 epoch 2025:

Code: [Select]

Vector([-115.22290285000018, 36.01385620910244, 717.0339638399593]) WGS 84
Vector([-115.22288898812235, 36.013852174830696, 745.6915649031728]) NAD83(2011) + NAVD88 height
Vector([-115.22288662114674, 36.013853610407665, 745.6749450693135]) NAD83(2011) + NAVD88 height_2025.1
Navd88 height changes some 1.5 cm from epoch 2020 to 2025.1...

The bench mark you have may have published NAVD88 elevations different from the geoid height given by the model...If you have some benchmarks you can share and I can look at them...

However, I downloades a combined (horizontal and vertical) control point from NGS and took the NAD83(2011) coordinates and converted to NAD83(2011) + NAVD88 height in MS and got following:

Code: [Select]

W 51 Lon -115.014397 Lat 36.067895 H 476.555 NAD83(2011)
W 51 Lon -115.014397 Lat 36.067895 H 504.846 NAD83(2011) + NAVD88 height

1    National Geodetic Survey, Retrieval Date = FEBRUARY  1, 2025 02:37:44 EST
GR1201 ***********************************************************************
GR1201  DESIGNATION -  W 51
GR1201                         *CURRENT SURVEY CONTROL
GR1201  ______________________________________________________________________
GR1201* NAD 83(2011) POSITION- 36 04 04.42112(N) 115 00 51.82751(W)   ADJUSTED
GR1201* NAD 83(2011) ELLIP HT-   476.555 (meters)        (06/27/12)   ADJUSTED
GR1201* NAD 83(2011) EPOCH   -  2010.00
GR1201* NAVD 88 ORTHO HEIGHT -   504.851 (meters)     1656.33  (feet) ADJUSTED
As you can see the transformed NAVD88 height is within 5 mm from published one...

[aggieair]

My benchmarks are actually Propeller AeroPoints that I'm using as check points. When they are converted to WGS84 / UTM + WGS84 EllipHT (using NOAA Vertical Datum Converter, one example attached), they are pretty matched up with my DSM exported from Agisoft in WGS84 / UTM + WGS84 EllipHT. This is the initial projection. I get only a ~7cm difference between my Aero check points and my DSM.

I think I am getting all the same DSM elevations because when I go to Export DEM after doing each Convert option, I am selecting NAD83(2011) / UTM Zone 12N + NAVD88 Geoid 12b from the drop down menu of the Export DEM window (see attached). I do the convert function just before this, but maybe me picking from the drop down menu is cancelling things out? If this is incorrect, how does one export a converted projection DEM?

[Paulo]

aggieair,

I think your problem is not configuring the transformation parameters using the Tools icon to the right of Coordinate System drop down menu. Check the following towgs84 parameters are entered  [-1.0053, 1.90921, 0.54157, 0.0267815, -0.00042078, 0.01093254, -0.00036891]. And you should get a good export DEM.

I tested with same point comparing VDATUM transfo to Metashape and got very similar results. As above 7 parameters concern ITRF14 realization of WGS84, at epoch 2010, I use the IGS14 or ITRF2014 target datum and specify 2010 as input and output epoch.

See following:

[aggieair]

I am not familiar with using commands in the Console window. Here I am trying some of the inputs you have in your example.

[Paulo]

I understand...

in API you have to define the variables nad83 and wgs84_12N as Coordinate Systems by using:

Code: [Select]

nad83 = Metashape.CoordinateSystem("EPSG::6318")
wgs84_12N = Metashape.CoordinateSystem('EPSG::32612')
by default the nad83 CS has the 7 TOWGS84 parameters defined as [0,0,0,0,0,0,0]

so in console you set the parameters to required  list by:

Code: [Select]

nad83.towgs84 = [-1.0053, 1.90921, 0.54157, 0.0267815, -0.00042078, 0.01093254, -0.00036891]
then define the coordinates of vector you want  to transform and do conversion from nad83 to wgs84_12N CS:

Code: [Select]

pt = Metashape.Vector((-111.886884,41.72064256,1335.827))
nad83.transform(pt,nad83,wgs84_12N)
Out[10]: 2025-02-05 15:11:02 Vector([426228.4682755239, 4619140.799984329, 1335.139839227436])
Quick class in API... if you need more orientation, please send me a PM.